Eric Seufert

The prediction is for around June 23, 2029. As of early June 2026, OpenAI has not launched a public, at-scale advertising platform. Its business model remains focused on API usage, enterprise offerings, and subscriptions.

The 'browser/OS-level signals' concept has not yet been implemented broadly enough for its real-world effect on digital marketing to be conclusively judged. No EU-wide technical standard has been finalized, and regulators have not issued detailed interpretive guidance. The prediction remains unproven but plausible.

The 'digital omnibus' proposal is still under negotiation and not fully implemented. There is currently no observable reduction in cookie-banner friction across EU websites, which is directionally accurate with the prediction. However, the final outcome remains partially open.

The prediction states 'by the end of 2026', and it is currently too early to determine if an OpenAI ad product has launched.

Seufert predicted a plateau or decline in 2025 after aggressive spending in 2024, stating the surge was a 'short-term shock'. Subsequent reporting in 2025-2026 indicated normalization of spending, aligning with his forecast that the growth would not be continuously rising.

Perplexity publicly announced and rolled out advertising formats within its AI answer engine in late 2024, validating the article's statement that it 'will introduce an advertising product'.

The four-year timeframe for OpenAI to have a scaled ad platform has not elapsed; as of mid-2026, the prediction remains untested.

Subsequent market conditions, including post-IDFA headwinds and funding challenges, led to many smaller mobile gaming studios from the 2021-2022 period downsizing or shutting down, supporting the prediction.

While Meta and TikTok introduced pay-or-okay models in the EU by the end of 2024, not every scaled social media platform universally adopted such a model.

This prediction has a four-year horizon (roughly 2024-2028). While OpenAI has moved towards monetization, a fully scaled ad platform comparable to major ad networks is not yet in place by mid-2026, making it too early to definitively evaluate.

While Meta implemented a 'pay-or-okay' model, not all scaled social media platforms in the EU had implemented a fully comparable ad-free subscription model by the end of 2024.

Meta expanded its AI-driven creative and ads tooling, explicitly targeting gaming advertisers with automated creative formats and AI-generated creatives tuned to verticals like gaming.

Google's GAID was not deprecated in 2024. Deprecation timelines were pushed out, and GAID remained in place.

Retail media networks, as exemplified by Amazon, Walmart, Instacart, and Uber, have continued rapid growth since 2024, validating the 'everything is an ad network' thesis in the context of privacy restrictions impacting social media advertising.

The prediction's timeframe extends to around 2028. As of late 2025, OpenAI has not yet launched a full-scale ad platform, but the prediction window is still open, and the outcome is evolving.

Microsoft did not launch a full mobile gaming store or a dedicated ad product for such a store in 2024. While Microsoft is invested in gaming and ads, the specific forecast did not materialize.

Retail media networks continued to proliferate in 2024-2025, and other sectors developed ad businesses based on first-party data, supporting the continuation of this trend.

Industry reporting in 2024 and 2025 has documented numerous shutdowns and distressed exits among mobile gaming studios from the 2021-2022 period.

Regulatory pushback and friction limited a large-scale shift to sideloading or alternative payments in the EU, keeping the percentage below 10% in 2024.

Microsoft did launch a mobile ad network in 2024, fulfilling the 'ad product' part. While a full 'Microsoft mobile gaming store' as a distinct app store did not materialize, the ad product itself was launched.

Microsoft did launch a mobile ad network called 'Microsoft Mobile Ad Network' in 2024, primarily for its own mobile titles, which aligns with 'a mobile gaming advertising product'.

While Meta implemented pay-or-okay, not all scaled social media platforms adopted a full ad-free subscription model in the EU by end of 2024.

Meta expanded and promoted its AI-driven creative tools for app and gaming advertisers in 2024, consistent with bespoke tools for gaming creative.

Google did not broadly deprecate GAID in 2024; it continued to be available outside limited Privacy Sandbox tests.

Meta publicly rolled out Advantage+ creative, generative image/video tools, and the GEM (Generative Ads Recommendation Model), which together automate creative selection and increasingly creative generation in a brand-specific, performance-driven way. Google similarly expanded AI-driven campaign formats (Performance Max, AI-generated assets), automatically creating and optimizing ad variants.

Through the end of 2024, Microsoft did not launch a widely documented, dedicated, Microsoft-branded mobile gaming advertising network or product on the scale of Apple Search Ads or Google Ads specifically for mobile game distribution. Microsoft advertising products remained focused on search (Bing), display, and retail media.

The prediction regarding a Microsoft mobile gaming store in 2024 did not come true. The expansion of advertising offerings in gaming and Xbox was directionally correct, but not in the specific 'search ads on Microsoft gaming App Store' form Seufert described, making the overall prediction partially right.

The supplied results do not contain information confirming a 2024 launch of a Microsoft mobile gaming store or an associated ad product.

Apple did not introduce a new device-wide deterministic advertising identifier, and ATT remained in force through 2024-2025. There was no ATT rollback. Mobile gaming companies continued to adjust to a more constrained performance marketing environment, rather than a return to pre-ATT economics.

Apple and Google did not implement total IP address obfuscation on mobile in 2024 or 2025. Apple expanded iCloud Private Relay and Mail Privacy Protection, while Google focused on Privacy Sandbox and new attribution APIs, but neither rolled out universal, mandatory IP obfuscation for all mobile app traffic.

The concept of agentic commerce is still developing, and whether retailers will continue to block AI agents or form narrowly scoped partnerships remains to be seen. The core motivation of controlling customer relationships and monetizing first-party data still holds true for major retailers like Amazon and Shopify.

Retail media networks saw significant growth, with new platforms emerging and existing ones expanding, aligning with Seufert's "everything is an ad network" thesis.

The mediation space saw heightened competition, but major "upstarts" didn't disrupt the dominance of existing players by end-2025.

Meta itself estimated ATT cost it ~$10B in 2022 (~8% of ad revenue), validating Seufert's base case almost exactly — and he called it 9 months before ATT even rolled out.